Evaluation of Diagnostic Tests Under Local Conditions

Based on the substantial variation in the results reported for the same groups of cattle by different investigators, it appears that there is a marked inconsistency in the application of criteria used for the interpretation of the test results.

Since the available diagnostic tests were developed in different settings with different breeds of cattle, it was deemed necessary to assess the accuracy of various diagnostic techniques under Cameroonian conditions. As an example, performing the CCT in Maroua and using ≥2 mm, ≥3 mm, and ≥ 4 mm as cut-off points and the presence of visible BTB lesion as a reference test, the estimated sensitivities were, respectively, 78.5%, 67.8%, and 57.1%, and the specificities were, respectively, 85.9%, 94.7%, and 96.5% (Awah-Ndukum et al. unpublished data). Using the SIT at cut-off points of ≥3 mm and ≥ 4 mm, the respective sensitivity was 82.1% and 71.4%, and the specificity was 91.2% and 96.5%. When the presence of TB lesion in addition to Ziehl-Neelsen smear microscopy was used as the reference test to define the disease status, a sensitivity, respectively, of 100%, 89.4%, and 73.6% and a specificity of 83.3%, 92.4%, and 93.9% at ≥2 mm, ≥3 mm, and ≥ 4 mm cut-off points were obtained. Furthermore, receiver operating characteristic (ROC) analysis showed better performance at the ≥3-mm cut-off (over 91%) compared to the ≥4- mm cut-off point (84%) suggesting that in Cameroon a stricter interpretation of the specific tuberculin skin tests would detect more BTB-positive cattle. Bronsvoort et al. (unpublished data) similarly obtained better values for the sensitivity and specificity of the CCT at a cut-off value of ≥3 mm compared to ≥4 mm.

The detection of lesions that are consistent with those caused by M. bovis during postmortal examination provides most of the information on which the prevalence and distribution of BTB in Cameroon are based. This technique too is flawed; there is ample evidence that under-recording and under-detection are very common, and the information provided is most likely a marked underestimation of the actual situation. Animals demonstrating poor health and diminished productivity are cus­tomarily the ones removed from herds and slaughtered for meat production. As they are usually disproportionately old-aged, abattoir inspection may not provide a true estimate of BTB in the local cattle population as the prevalence of BTB is expected to be higher in this group of animals.

Interpretation and the detection of TB-like lesions in slaughtered cattle can be difficult. Routine meat inspection generally only detects lesions in about 50% of carcasses containing tuberculous lesions. These lesions may resemble abscesses (with yellowish pus) or manifest as firm, yellowish, nodular lesions (often “gritty” on cutting), and are commonly detected in the lungs and associated lymph nodes (over 60%), followed by lymph nodes of the head, mesenteric lymph nodes, and the liver. The granulomatous lesions may easily be confused with parasitic granulomas, non-specific inflammatory reactions (Corner 1994; Shitaye et al. 2006; Edwards et al. 1997), and lesions caused by Nocardia, Corynebacterium, and other pyogranuloma-causing organisms (Grist 2008), and wrongly diagnosed as tubercu­lous unless confirmed by culture. Nonetheless, in slaughtered cattle, TB lesions were 3-5 times more prevalent than similar lesions caused by a different etiological agent.

12.4